One of my disks in a raidz2 array crashed a few days ago and it was time to hotswap a disk. zpool status showed a faulted drive:
raidz2 DEGRADED 0 0 0 c3t6d0 ONLINE 0 0 0 c3t7d0 FAULTED 27 85.2K 0 too many errors c5t0d0 ONLINE 0 0 0 c5t1d0 ONLINE 0 0 0
The disk is attached into an LSI Logic SAS1068E B3 SATA card which has eight SATA ports. I used lsiutil to find out that there were indeed some serious problems with the disk:
Adapter Phy 7: Link Up Invalid DWord Count 306,006 Running Disparity Error Count 309,292 Loss of DWord Synch Count 0 Phy Reset Problem Count 0
I’m not sure what “Invalid DWord Count” and “Running Disparity Error Count” means, but that indeed doesn’t look good. I guess I need to do some googling after this. zpool status showed problems with disk c3t7d0 which is mapped into the 8th disk in the LSI card.
I replaced the old disk and added the new disk into the system on the fly. The LSI card noticed and initialized the disk, but with a different id. The disk is now c3t8d0. This is propably because the disk wasn’t the same. I ordered zfs to replace the old disk with the new one with command “zpool replace tank c3t7d0 c3t8d0”
raidz2 DEGRADED 0 0 0 c3t6d0 ONLINE 0 0 0 14.2M resilvered replacing DEGRADED 0 0 0 c3t7d0 FAULTED 27 89.2K 0 too many errors c3t8d0 ONLINE 0 0 0 2.03G resilvered c5t0d0 ONLINE 0 0 0 14.2M resilvered c5t1d0 ONLINE 0 0 0 13.5M resilvered
That’s it. The resilver took me 3h16m to complete.